The immunogenicity, tolerability and interchangeability of two hepatitis A vaccines, Vaqta (Merck and Co.) and Havrix (SmithKline) were studied in a randomized, crossover, controlled clinical trial. Vaccine was administered to 201 volunteers at 0 and 26 weeks in one of four vaccine regimens: Havrix-Havrix; Havrix-Vaqta; Vaqta-Havrix or Vaqta-Vaqta. Seroconversion rates (>/=10 mIU/ml) for those whose first dose was Vaqta or Havrix, respectively, were: 41/96 (43%) versus 30/95 (32%) (P=0.15) at 2 weeks and 91/98 (93%) versus 84/97 (87%) (P=0.43) at 4 weeks, and 100% at 26 weeks. Geometric mean concentrations (GMC) of total antibody to hepatitis A virus (anti-HAV) for Vaqta and Havrix were 189 and 114 mIU/ml (P=0.011) at 4 weeks and 234 and 136 mIU/ml (P<0.001) at 26 weeks. At 30 weeks, the GMC after two doses of Havrix was 2612 mIU/ml compared with 5497 after two doses of Vaqta (P<0.001). The GMC in the Havrix-Vaqta group was 5672 mIU/ml compared with 3077 mIU/ml in the Vaqta-Havrix group (P<0.001). Less than half of vaccine recipients reported tenderness or pain. In this study, seroconversion rates of the two vaccines were similar. Vaqta produces significantly higher anti-HAV antibody than Havrix. Crossover immunization is well tolerated and results in high antibody concentrations, especially when Vaqta is the booster dose. The significance of higher anti-HAV antibody concentrations in terms of long-term protection is unknown.